Search for quasi-periodic signals in magnetar giant flares

TL;DR

This study uses a Bayesian method to analyze magnetar giant flare data, finding new potential oscillation signals that could inform models of neutron star properties and magnetic field strengths.

Contribution

It applies a noise-aware Bayesian inference method to search for quasi-periodic signals in magnetar flares, challenging previous detections and proposing new candidate oscillations.

Findings

No confirmation of previously reported high-frequency lines.

Identification of two new potential oscillation candidates at 9.2 Hz and 7.7 Hz.

Implications for weaker magnetic fields and lower crust shear velocities in magnetars.

Abstract

Quasi-periodic oscillations (QPOs) discovered in the decaying tails of giant flares of magnetars are believed to be torsional oscillations of neutron stars. These QPOs have a high potential to constrain properties of high-density matter. In search for quasi-periodic signals, we study the light curves of the giant flares of SGR 1806-20 and SGR 1900+14, with a non-parametric Bayesian signal inference method called D$^3$PO. The D$^3$PO algorithm models the raw photon counts as a continuous flux and takes the Poissonian shot noise as well as all instrument effects into account. It reconstructs the logarithmic flux and its power spectrum from the data. Using this fully noise-aware method, we do not confirm previously reported frequency lines at $\nu\gtrsim17\,$Hz because they fall into the noise-dominated regime. However, we find two new potential candidates for oscillations at $9.2\,$Hz (SGR 1806-20) and $7.7\,$Hz (SGR 1900+14). If these are real and the fundamental magneto-elastic oscillations of the magnetars, current theoretical models would favour relatively weak magnetic fields $\bar B\sim 6\times10^{13} - 3\times10^{14}\,$G (SGR 1806-20) and a relatively low shear velocity inside the crust compared to previous findings.